建筑垃圾再生料影像特征及含水率识别方法

针对目前常规方法难以准确实时监测建筑垃圾再生料含水率的问题，通过对建筑垃圾再生料宏细观形态及色彩特征进行定性分析，基于MATLAB提取物料数字影像特征值与含水率进行相关性分析，结果表明，除色调外，建筑垃圾再生料数字影像特征参数与含水率的相关性显著；采用步进法对特征参数与含水率进行线性回归分析，拟合出建筑垃圾再生料含水率计算模型，并对其进行试验及现场验证，检验结果表明，该建筑垃圾再生料含水率计算模型在实验室条件下误差为-0.2%~0.9%，现场条件下为-3.0%~2.6%，可为以建筑垃圾再生料为原料的自动化制浆系统建设及充填过程控制提供依据。

A method for moisture content identification of construction waste recycled materials based on digital image features

Aiming at the problem that current conventional methods are difficult to accurately and real-time monitor the moisture content of construction waste recycled materials, through qualitative analysis of the macro and micro morphology and color characteristics of construction waste recycled materials, based on Matlab to extract the digital image characteristic value of materials and the correlation analysis of moisture content, the results show that, in addition to color,The correlation between the digital image characteristic parameters and water content of construction waste recycled material is significant.Linear regression analysis of characteristic parameters and water content was carried out by step method to fit the calculation model of water content of construction waste recycled material, and the test and field verification were carried out. The test results showed that the error of the calculation model of water content of construction waste recycled material was -0.2%~0.9% under laboratory conditions, and -3.0%~2.6% under field conditions.It can provide a method basis for the construction of automatic pulping system and the control of filling process based on construction waste recycling materials.